基于BPNN的超声波电动机转速控制与实验研究

Rotating-Speed Control and Experimental Research of Ultrasonic Motor Based on BPNN

为改善行波型超声波电动机因强非线性问题引起的转速跟踪准确性低和控制平稳性差等问题,设计了基于BP神经网络(BPNN)超声波电动机转速控制算法,并对设计算法进行了仿真分析;通过超声波电动机实验系统证明了该控制算法的有效性。研究结果表明,基于BPNN的控制策略可实现超声波电动机快速响应和高精度速度跟踪,控制误差小于2r/min。

Ultrasonic motor is difficult to achieve an ideal stability and precision due to the nonlinear characteristics.In order to improve control precision and stability,a BPNN(BP neural network)-based control strategy was proposed to control the rotating-speed of the ultrasonic motor.After some simulations,an experimental system was setup and some experiments were performed to testify the control algorithm.The experimental results show that the BPNN-based control strategy can achieve rapid response and high precision speed tracking;the control error is less than 2 r/min.